BACKGROUND: Mortality rates due to pulmonary embolism (PE) are difficult to estimate often due to the presence of comorbid disease. OBJECTIVES: To determine the accuracy of hospital records in identifying PE cases, PE-related mortality, and the number of PE-related deaths which are potentially preventable. METHODS: Retrospective chart review of PE cases hospitalized at The Ottawa Hospital over an 8 year period. Cases were reviewed to determine accuracy of coding, as well as the certainty with which PE was the cause of death. In PE-related deaths, a determination was made as to whether any interventions may have been life-saving. RESULTS: 498 cases of 612 (81%) cases coded as PE were correctly coded. 111 (22%) died during hospitalization, 63% of deaths were attributed to PE. The presence of a cardiorespiratory comorbidity or cancer was independently associated with an increased rate of death due to PE. 54% of PE-related deaths were determined to be potentially preventable, most commonly by appropriate DVT prophylaxis. A significantly higher number of cancer patients as compared to non-cancer patients may have potentially had their death due to PE prevented by an inferior vena cava filter (IVCF). Systemic thrombolysis was deemed to be potentially life-saving in 1/38 PE-related deaths. CONCLUSION: Hospital mortality due to clinically recognized PE can be determined by chart review of PE cases identified using the ICD coding system. Death due to PE is often potentially preventable; in the subgroup with cancer and DVT/PE, an IVCF may be a potentially useful intervention to prevent death due to PE. Prospective studies are needed to confirm these results. Copyright 2009 Elsevier Ltd. All rights reserved.
BACKGROUND: Mortality rates due to pulmonary embolism (PE) are difficult to estimate often due to the presence of comorbid disease. OBJECTIVES: To determine the accuracy of hospital records in identifying PE cases, PE-related mortality, and the number of PE-related deaths which are potentially preventable. METHODS: Retrospective chart review of PE cases hospitalized at The Ottawa Hospital over an 8 year period. Cases were reviewed to determine accuracy of coding, as well as the certainty with which PE was the cause of death. In PE-related deaths, a determination was made as to whether any interventions may have been life-saving. RESULTS: 498 cases of 612 (81%) cases coded as PE were correctly coded. 111 (22%) died during hospitalization, 63% of deaths were attributed to PE. The presence of a cardiorespiratory comorbidity or cancer was independently associated with an increased rate of death due to PE. 54% of PE-related deaths were determined to be potentially preventable, most commonly by appropriate DVT prophylaxis. A significantly higher number of cancerpatients as compared to non-cancerpatients may have potentially had their death due to PE prevented by an inferior vena cava filter (IVCF). Systemic thrombolysis was deemed to be potentially life-saving in 1/38 PE-related deaths. CONCLUSION: Hospital mortality due to clinically recognized PE can be determined by chart review of PE cases identified using the ICD coding system. Death due to PE is often potentially preventable; in the subgroup with cancer and DVT/PE, an IVCF may be a potentially useful intervention to prevent death due to PE. Prospective studies are needed to confirm these results. Copyright 2009 Elsevier Ltd. All rights reserved.